Andrew Holtz is one of our longtime contributors and a past president of the Association of Health Care Journalists. He tweets at @HoltzReport.
We need people like genetics pioneer Craig Venter. But as with most pioneers, it’s generally a good idea to follow them from a safe distance.
Venter recently announced that he underwent surgery after genetic testing he helped develop led to a discovery of what he termed “high-grade” prostate cancer. Venter says there were no other indications of cancer, such as an elevated PSA test result. It sounds like a straightforward sequence of events: Innovative test finds something, follow-up testing confirms it, treatment deals with it. But with screening tests, nothing is straightforward, especially not the well-intended, but often misguided, belief that finding cancer earlier makes it easier to cure.
There are a lot of things we don’t know about the testing that Venter underwent. What we do know: His Health Nucleus business advertises a $25,000 workup that includes genetic screening, imaging and other tests and consultations. The article about his cancer announcement reported that “Venter said Wednesday that 40 percent of the Health Nucleus clients, who buy the service believing they are healthy, are found to have “serious disease,” and 20 percent have life-threatening disease.” Actually, that sounds on the low side. Really. Even just for prostate cancer. A look at autopsy studies indicates that more than a third of men about Venter’s age, 70, had undiagnosed prostate cancer when they died. (One study reported that 36 percent of Caucasian men who died in their 70s were found to have a tumor in their prostate. A systematic review of autopsy studies estimated that a third to a half of men who die in their 70s have prostate cancer.)
Lots of uncertainty surrounds how best to screen, treat prostate cancer
So finding prostate cancer in an older man is no great trick. The tricky thing is deciding what to do about it. There’s a big difference between chasing down symptoms that reveal a problem and just stumbling on something that looks like a problem. Venter’s experience might be comparable to that of a man learning he has cancer in his prostate after getting surgery to treat an enlarged prostate. Benign prostatic hyperplasia is very common as men age, the main issue being difficulty urinating. When the tissue removed from the prostate is examined, sometimes cancer is found. Sometimes these men are advised to “watch and wait” (perhaps by having regular PSA tests) and sometimes they are treated, depending on how threatening the tumor appears to be. Venter reportedly said his cancer was “high-grade.” We don’t know exactly what that means, but his decision to get treatment sounds like what many men would do after learning they have “incidental” prostate tumors (that is, the tumors were discovered incidental to other treatment or tests).
It’s no surprise that when a tumor is found that appears to be “high-grade,” many men want it taken out or zapped with radiation. But be careful not to declare that the surgery or radiation “saved” the life of any individual patient, an issue we recently raised with actor Ben Stiller’s prostate cancer essay. The PSA test has revealed a lot of prostate cancers that would not have been noticed otherwise, and while prostate cancer death rates have been declining, as a National Cancer Institute summary put it, “associations with screening patterns are inconsistent. The observed trends may be due to screening, or to other factors such as improved treatment. And as one study after another has failed to find convincing evidence that testing healthy men was clearly beneficial to them (especially when you take into account the impotence, incontinence and other harms of treatment), medical societies and expert panels have been dialing back their prostate cancer screening recommendations.
That doesn’t mean prostate cancer treatments are ineffective. It’s just complicated. Prostate cancer tends to grow very, very slowly; so slowly that most older men with cancer would probably die of something else before the tumor grew and spread enough to kill them. And even when prostate cancer is treated and seems to be completely gone, it sometimes returns.
Since there is a great deal of debate about the effectiveness of prostate cancer treatment in men diagnosed in the usual ways, even after decades of study involving data on multitudes of patients, it is certainly impossible to say whether Venter’s treatment following a highly unusual diagnosis will make an important difference.
Early diagnosis is not always better–sometimes it’s worse
You hear all the time that finding something earlier is better. Sorry, but it ain’t always true. Indeed, sometimes it can be worse. How could it be worse? If the timing of the treatment doesn’t make an important difference.
One prominent example is treatment of diabetes and the growing discussion around “pre-diabetes.” Diabetes is serious–not only can a blood sugar crisis kill, it steadily damages nerves, blood vessels and other organs, leading to heart attacks, amputations and on and on. It seems logical that starting treatment earlier would help stave off those threats. But as we’ve discussed and as a recent analysis published in BMJ laments, the “epidemic of pre-diabetes” should be scrutinized. If poor test results help motivate someone to lose weight and get off the couch, great. But turning them into “patients” sooner–putting them on drugs sooner, making them go to the doctor more, giving them more tests, hitting them with all the bills–that makes sense only if there are benefits to beginning treatment before someone develops diabetes, and so far there isn’t much evidence there are any.
And then there are conditions, such as Alzheimer’s, for which no effective treatment exists. Predictive tests are being developed, and some people may want to have a personalized probability estimate for estate planning or other purposes, but many others prefer to simply live their lives, given the lack of options around treatment.
Prostate cancer is sort of a muddy in-between example: There are treatments, they can affect the course of the disease, but whether they are a net benefit or not to any specific individual is almost impossible to say.
How “lead-time bias” can create an illusion of success
Here’s where we talk about moving the starting line, or in medical jargon, the “lead-time bias.” Cancer treatment is often assessed using 5-year survival rates; that is, what percentage of patients are alive five years after diagnosis. That benchmark is a useful way to compare treatments. But it is easy to misinterpret in ways that create an illusion of success.
Here’s how things can go awry. Let’s say you have a disease that typically kills patients two years after obvious symptoms appear and they don’t have any treatment. And then say you have a treatment that extends their lives by an additional year, so that patients who start treatment soon after symptoms typically live for three years. Then researchers come up with a blood test that reliably detects the disease three years before symptoms appear.
Naturally, doctors start offering treatment earlier. Researchers then do a study and determine that patients are now living six years after diagnosis. Headlines blare: “Patients Living Twice as Long after Early Diagnosis.” After all, patient survival went from three years to six years. As some of you have probably noticed, earlier diagnosis did not actually improve survival at all. It just moved the starting line three years earlier.
Really, in this harsh scenario, earlier diagnosis made things worse. Instead of going through three years of treatment, those diagnosed early went through six years of treatment, with all the side effects and costs, and six years of having the terminal diagnosis hanging over them and their families. Before the early diagnosis test, they spent those first three years believing they were healthy (just like the description of Venter’s Health Nucleus clients). In effect, the early diagnosis robbed them of three years of normal life.
Enthusiastic promotion should not be confused with medical evidence
In the real world things are rarely that clear cut. Yet the scenario is not entirely fiction. There are a number of companies that make their money giving people tests that do not clearly help people live longer or better. Some of them send fancy motor homes around the country offering ultrasound tests for abdominal aortic aneurysms and other artery irregularities, even though there’s no evidence such screenings can help an average person live any longer. Though the popularity of whole-body CT scanning seems to have faded somewhat, it was long touted as a way to find things early. Indeed, those CT scans do find things, but on balance people are more likely to be harmed (or not helped) by the screening.
Like many pioneers, Craig Venter is passionate. There is no reason to doubt the sincerity of his belief that someday his Health Nucleus testing or later iterations will improve human health. But it is not healthy to accept passionate belief without scrutiny. Nor should enthusiastic promotion be confused with medical evidence. Reporters covering Craig Venter’s promotion of Health Nucleus have asked about the apparent futility of whole-body imaging in the past. His response: The technology is better now. Certainly the technology does produce higher resolution images, but pixel count is not proof of benefit. Until the Health Nucleus testing sequence is itself tested, it is not known whether the results improve the quality or length of life–or just turn apparently healthy people into patients, turning normal life of work and play, family, friends and passions into a life shadowed by tests and treatments, doctors and clinics.
Update: The original post has been edited to note that prostate cancer deaths have been declining, but that it is not clear how much credit goes to screening.